AC-RT66U completely unreliable with or without merlin

[timredfern]

This is driving me up the wall.

I bought the AC-RT66U to set up load balancing in a small office.

I've tried the official firmware and various versions of Merlin including the latest.

I'm seeing the following:

If I turn on dual WAN (in either mode) the service bombs. Clients (both wired and wireless) lose internet connectivity constantly (although not actually disconnected). I also see crazy variable latency especially on 5Ghz wireless, of up to 2 seconds.

If I go back to single WAN the service is more reliable but still has 'episodes' where clients lose connectivity. Latency is still very variable: on 5Ghz for example I'm seeing this:

64 bytes from 173.194.112.63: icmp_seq=1138 ttl=53 time=32.804 ms
64 bytes from 173.194.112.63: icmp_seq=1139 ttl=53 time=38.706 ms
64 bytes from 173.194.112.63: icmp_seq=1140 ttl=53 time=32.930 ms
64 bytes from 173.194.112.63: icmp_seq=1141 ttl=53 time=35.190 ms
64 bytes from 173.194.112.63: icmp_seq=1142 ttl=53 time=34.503 ms
64 bytes from 173.194.112.63: icmp_seq=1143 ttl=53 time=50.532 ms
64 bytes from 173.194.112.63: icmp_seq=1144 ttl=53 time=100.826 ms
64 bytes from 173.194.112.63: icmp_seq=1145 ttl=53 time=147.947 ms
64 bytes from 173.194.112.63: icmp_seq=1146 ttl=53 time=188.270 ms
64 bytes from 173.194.112.63: icmp_seq=1147 ttl=53 time=230.915 ms
64 bytes from 173.194.112.63: icmp_seq=1148 ttl=53 time=247.214 ms
64 bytes from 173.194.112.63: icmp_seq=1149 ttl=53 time=33.361 ms
64 bytes from 173.194.112.63: icmp_seq=1150 ttl=53 time=33.834 ms
64 bytes from 173.194.112.63: icmp_seq=1151 ttl=53 time=35.470 ms
64 bytes from 173.194.112.63: icmp_seq=1152 ttl=53 time=45.432 ms
64 bytes from 173.194.112.63: icmp_seq=1153 ttl=53 time=37.091 ms
64 bytes from 173.194.112.63: icmp_seq=1154 ttl=53 time=32.407 ms
64 bytes from 173.194.112.63: icmp_seq=1155 ttl=53 time=70.876 ms
64 bytes from 173.194.112.63: icmp_seq=1156 ttl=53 time=113.718 ms
64 bytes from 173.194.112.63: icmp_seq=1157 ttl=53 time=156.980 ms
64 bytes from 173.194.112.63: icmp_seq=1158 ttl=53 time=208.234 ms
64 bytes from 173.194.112.63: icmp_seq=1159 ttl=53 time=252.142 ms
64 bytes from 173.194.112.63: icmp_seq=1160 ttl=53 time=35.845 ms
64 bytes from 173.194.112.63: icmp_seq=1161 ttl=53 time=54.509 ms
64 bytes from 173.194.112.63: icmp_seq=1162 ttl=53 time=59.040 ms
64 bytes from 173.194.112.63: icmp_seq=1163 ttl=53 time=42.374 ms
64 bytes from 173.194.112.63: icmp_seq=1164 ttl=53 time=32.967 ms
64 bytes from 173.194.112.63: icmp_seq=1165 ttl=53 time=33.778 ms
64 bytes from 173.194.112.63: icmp_seq=1166 ttl=53 time=39.814 ms
64 bytes from 173.194.112.63: icmp_seq=1167 ttl=53 time=91.781 ms
64 bytes from 173.194.112.63: icmp_seq=1168 ttl=53 time=134.492 ms
64 bytes from 173.194.112.63: icmp_seq=1169 ttl=53 time=179.720 ms
64 bytes from 173.194.112.63: icmp_seq=1170 ttl=53 time=243.224 ms
64 bytes from 173.194.112.63: icmp_seq=1171 ttl=53 time=257.566 ms

When I connect directly to DSL by ethernet the latency is solid 7ms. What the hell is the ASUS router doing?

The ASUS logs show clients are contantly performing DHCP requests:

Nov 17 13:11:43 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.2.131 18:ee:69:0a:db:16
Nov 17 13:11:43 dnsmasq-dhcp[290]: DHCPACK(br0) 192.168.2.131 18:ee:69:0a:db:16 Apple-TV-2
Nov 17 13:12:55 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.2.131 18:ee:69:0a:db:16
Nov 17 13:12:55 dnsmasq-dhcp[290]: DHCPACK(br0) 192.168.2.131 18:ee:69:0a:db:16 Apple-TV-2
Nov 17 13:14:07 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.2.131 18:ee:69:0a:db:16
Nov 17 13:14:07 dnsmasq-dhcp[290]: DHCPACK(br0) 192.168.2.131 18:ee:69:0a:db:16 Apple-TV-2
Nov 17 13:14:55 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.2.131 18:ee:69:0a:db:16
Nov 17 13:14:55 dnsmasq-dhcp[290]: DHCPACK(br0) 192.168.2.131 18:ee:69:0a:db:16 Apple-TV-2
Nov 17 13:16:07 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.2.131 18:ee:69:0a:db:16
Nov 17 13:16:07 dnsmasq-dhcp[290]: DHCPACK(br0) 192.168.2.131 18:ee:69:0a:db:16 Apple-TV-2

When Dual WAN is active I seem to be seeing errors like this:

Nov 17 11:04:08 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.43.58 24:24:0e:c1:55:31
Nov 17 11:04:08 dnsmasq-dhcp[290]: DHCPNAK(br0) 192.168.43.58 24:24:0e:c1:55:31 wrong network

The router is not overloaded. The largest number of clients it has had is 21, at the moment cpu usage is 2-3% and 27% memory used.

Has anyone used the AC-RT66U sucessfully with or without dual WAN? I'm about to throw it out.

Tim Redfern

 

[L&LD]

http://www.snbforums.com/threads/no...l-and-manual-configuration.27115/#post-205573

Did you follow the (important) steps in the link above after you flashed new firmware to test with?

If you flashed, reset, configured and tested properly and you are still getting the same issues then I would agree to get a newer model.

Many RT-AC66U's I've seen were broken just after a year of use for various customers.

 

[john9527]

The ASUS logs show clients are contantly performing DHCP requests:

Nov 17 13:11:43 dnsmasq-dhcp[290]: DHCPREQUEST(br0) 192.168.2.131 18:ee:69:0a:db:16
Nov 17 13:11:43 dnsmasq-dhcp[290]: DHCPACK(br0) 192.168.2.131 18:ee:69:0a:db:16 Apple-TV-2

Known bug with Apple TV, not the router. You need to either set a static ip or switch to a wired connection.

 

[Rooter]

Has anyone used the AC-RT66U sucessfully with or without dual WAN? I'm about to throw it out.

Tim Redfern

While on Merlin's firmware I've used Dual WAN on an RT-AC66U with success (and no significant issues at the time other than IPv6), over both wired and wireless, but this was over a year ago. I've since upgraded to a newer model.